Draft 8

Single Laboratory Validation Acceptance Criteria

(Chemistry methods)
Purpose:
1. To provide guidance to method submitters on the criteria to be used for accepting
the method as an SLV method.
2. To provide uniformity in the review of methods submitted to AOAC.
3. To assist those laboratories engaged in single lab validation in understanding the
types of information the AOAC requires.
Acceptance Criteria:
1. Is the method clearly and understandably written? (Where alternatives are permitted,
are the circumstances for each option clearly and explicitly defined?)
2. Is the scope and applicability clearly stated? Methods submitted to the Dietary
Substances Project must be applicable to one or more of the following matrices or
commercial products:
2.1 The parent botanical (vouchered where possible as authentic)
2.2 A negative control or blank similar to the parent botanical (vouchered if
possible)
2.3 Concentrated powders and extracts
2.4 Tablets/pills and capsules (gel caps - hard or soft)
2.5 Combination products
(Describe steps taken to authenticate reference botanicals, negative controls and
marketed products, e.g. exhaustive extraction, certificate of analysis by a certified
laboratory, use of tracers, etc.)
3. Has the method been "optimized?" Have efforts been made to decrease run times,
increase resolution, improve peak shapes, minimize extraction time and improve
extraction efficiency, separations, selectivity, and address stability. Have tests for
analyte stability during processing and under typical conditions of storage been
addressed where appropriate?
4

Has provision been made for confirmation of analyte identity? (The method may be
sufficiently specific so that confirmation of identity may not be necessary.)

Was ruggedness testing done during the method development phase identifying
variables tested. Have these critical control points been written into the method?
(Note: ruggedness delineates the limits of minor variables; optimization applies to
major variables.)

Are the following performance characteristics available?

6.1 The recommended analytical range is LOQ to 200% of expected analyte
concentration.
6.2 Does calibration cover the analytical range? At least 5 concentrations should
be used for calibration. Are reference and internal standards available and is
the source identified? Are reference standards stable as prepared for use?
(The stability of the analytical curve should also be established by duplicating
it on another day. Note: It has been repeatedly pointed out that a linear
calibration curve is not a critical item, but merely a convenience, and r values
close to one may be deceptive [cf. Analyst 113, 1469 (1968)]. The

calibration curves for immunoassays are negative [0 concentration is a

maximum] and exponential.)
6.3 Accuracy/Recovery: Was recovery determined from spiked blanks or
samples with at least 7 independent analyses per concentration level at a
minimum of 3 concentration levels covering the analytical range.
Independent means at least at different times. If no confirmed (natural) blank
is available, the average inherent (naturally containing) level of the analyte
should be determined on at least 7 independent replicates. Was accuracy
determined using a Standard Reference Material from NIST if available or
other certified reference material; or if possible, by reference to an internal
standard. (It is critical during method development to establish a reference
point to which all assays can be referred. When this material is ultimately
assayed, all referred assays can be adjusted accordingly.
a. Precision (repeatability standard deviation): Were the number and types of
analyses sufficient for obtaining a good estimate of accuracy?
Typically for single lab validation one performs r replicate analyses of m test
portions over a period of d days for each sample type (matrix) n, where r is
the number of replicates (2,3 ), m is the number of test portions in each
group, d is the number of days, and n is the number of different sample
types.
r x m should never be less than 10
n should be at least 2, preferably more
d should be at least 2
The calculated HORRAT value should lie between 0.3-1.3. The formula
must be given to show that one is calculating the within-lab value (see
attached).
Recommended recovery and precision limits for single lab validation:
Concentration
Repeatability (%)
Recovery (%)
100%
1
98-101
10%
1.5
95-102
1%
2
92-105
0.1%
3
90-108
0.01%
4
85-110
6
80-115
10g/g (10 ppm)
8
75-120
1 g/g (1 ppm)
10 ng/g(10 ppb)
15
70-125
6.

Is method freely available (no restrictions on use and no copyright restrictions)?

AOAC will acknowledge the method source.

(For the use of the AOAC staff, the above information should be presented generally
following the format found in the "Guidelines for Single Laboratory Validation of Chemical
Methods for Dietary Supplements and Botanicals," page 33-34, and outlined below.)
Recommended Format for SLV Study Report
1. Method identification: title, authors, contact information, published references
2. Applicability :scope, abstract clearly identifying the analyte(s), matrix(ices), analytical range,
safety
3. Principle: may be included in abstract
4. Definitions (units if needed)
5. Reagents/supplies: include reference standards, calibration standards, etc.
6. Apparatus: include instruments and equipment used

7. Sampling: types of samples, amounts, sample handling and storage, stability, preparation of
test samples
8. Method: include calibration and procedure and include or refer to #1
9. Calculations summary: detection/quantitation limits, recovery, precision, and additional
information such as stability and measurement uncertainty (2RSDr)
10. Other pertinent information; e.g. confirmation of analyte identity and ruggedness (may be
placed in #8)
11. Conclusions